Spatial-Temporal Evidence Combination Method for Multi-sensor Target Recognition Based on Prospect Theory
Zihan Zhang, Zhentao Hu, Junwei Li, Yujie Su · 2022
Evidence theory has been widely used in the field of spatial-temporal information fusion, and since target recognition information fusion exhibits obvious dynamic and sequential characteristics. In order to realize the target recognition information fusion, spatial-temporal evidence is combined based on prospect theory and credibility decay model. Firstly, introducing the prospect value function to evaluate and adjust for evidence credibility, after that, an weight for each spatial evidence can be obtained to achieved the spatial information fusion. Finally, evidence combining by evidence discount criterion and Dempster combination rule to achieve sequential combination of temporal evidence. In this way, judgments for updated data of target recognition are data-driven, and makes more reasonable. The numerical simulation indicate that the proposed method fully reflects the dynamic and sequential characteristics of spatial-temporal evidence combination. Moreover, it has a strong capability to handle conflict information.